Defective innate immunity predisposes murine neonates to poor sepsis outcome but is reversed by TLR agonists

Abstract

Neonates exhibit an increased risk of sepsis mortality compared with adults. We show that in contrast to adults, survival from polymicrobial sepsis in murine neonates does not depend on an intact adaptive immune system and is not improved by T cell-directed adaptive immunotherapy. Furthermore, neonates manifest an attenuated inflammatory and innate response to sepsis, and have functional defects in their peritoneal CD11b(+) cells. Activation of innate immunity with either a Toll-like receptor 4 (TLR4) or TLR7/8 agonist, but not a TLR3 agonist, increased the magnitude, but abbreviated the early systemic inflammatory response, reduced bacteremia, and improved survival to polymicrobial sepsis. TLR4 agonist pretreatment enhanced peritoneal neutrophil recruitment with increased oxidative burst production, whereas the TLR7/8 agonist also enhanced peritoneal neutrophil recruitment with increased phagocytic ability. These benefits were independent of the adaptive immune system and type I interferon signaling. Improving innate immune function with select TLR agonists may be a useful strategy to prevent neonatal sepsis mortality.

Figures

Figure 1

Neonatal mice demonstrate altered peritoneal cell content and function compared with adult mice. Percentages (A) and phagocytic activity (B) (Dil+) of B220+CD11b+ peritoneal B1 cells and GR-1+Ly6ClowCD11bhigh neutrophils (PMN) from neonates versus young adults. Percentage of Dil+ cells refers to percentage of that cell population that registered greater than the baseline MFI (expressed as greater than 1000 MFI). Data shown are representative of 4 separate identical experiments. Bar graphs represent means with error bars representing standard deviations (*P < .05 by Student t test).

Figure 2

Adaptive immunity does not contribute to neonatal sepsis survival. Kaplan-Meier sepsis survival curves. (A) Sepsis survival in C57BL/6 and RAG-1–deficient adult mice following administration of cecal slurry (1 mg/g LD10). Adult sepsis survival in RAG-1–null mice (■) was significantly less (36%; P < .05 by Fisher exact test) than C57BL/6 wild-type adults (□; 100%). (B) Sepsis survival in C57BL/6 and RAG-1–deficient neonatal mice following administration of cecal slurry (1.3 mg/g LD70). Neonatal sepsis survival was not statistically different in C57BL/6 wild-type (♦) or RAG-1–deficient (◇) mice. (C) Sepsis survival in wild-type C57BL/6 neonates following anti-GITR (■) or isotype control (□) pretreatment followed by administration of cecal slurry (1.3 mg/g LD70). Neonatal sepsis survival was not statistically different in anti-GITR (■) or isotype control (□) pretreatment groups.

Figure 3

Pretreatment with TLR4 (LPS) or TLR7/8 (resiquimod) agonists enhances survival of neonatal mice with polymicrobial sepsis. Kaplan-Meier survival curve for neonates following TLR agonist pretreatment and subsequent administration of an approximate LD70 quantity of cecal slurry (1.3 mg/g LD70). LPS (●), resiquimod ([diao), poly I:C (□), and sham (normal saline; ▵). Survival was significantly (*P < .05 by Fisher exact test) improved in the LPS (61%)– and resiquimod (56%)–treated groups as compared with sham (35%) and poly I:C (33%).

Figure 4

TLR4 (LPS) and TLR7/8 (resiquimod) pretreatment reduces bacteremia during polymicrobial sepsis. Bacteremia (log of colony forming units/mL) in sham (normal saline, □) and TLR agonist–(resiquimod-, ; LPS-, ▨; poly I:C–, ) pretreated neonates (n = 7) at 6 and 12 hours after sepsis. Medians, are indicated by ▬ and quartiles (25th and 75th percentiles) by the surrounding rectangles. Statistical significance (*P < .05 by 1-way ANOVA) was present for resiquimod versus sham (normal saline), resiquimod versus poly I:C, LPS versus sham, and LPS versus poly I:C at 6 hours after sepsis as well as LPS versus resiquimod, sham, and poly I:C at 12 hours after sepsis.

Figure 5

TLR pretreatment increases and abbreviates the early neonatal inflammatory response following sepsis. Plasma cytokine levels and chemokine levels (pg/mL) at 4 and 24 hours after TLR agonist treatment (time points −20 and 0 hours, respectively) and at 12 and 24 hours after sepsis. LPS (●), resiquimod (◇), poly I:C (□), and sham (normal saline; ▵). The time point “zero” refers to the time sepsis was initiated. Values plotted represent medians, with error bars representing quartiles (75%/25%). Values that reached statistical significance (P < .05 by 1-way ANOVA) were too numerous to represent and are reported in “Results.”

Figure 6

TLR7/8 agonist (resiquimod) enhances peritoneal B1 and neutrophil recruitment and phagocytosis, whereas TLR4 agonist (LPS) enhances neutrophil recruitment and ROS production. (A) Percentages of neonatal peritoneal neutrophils (PMN) and peritoneal B1 cells from nonseptic sham (normal saline)–pretreated versus poly I:C–pretreated, resiquimod-pretreated, and LPS-pretreated neonatal mice at 24 hours after pretreatment. (B) Percentage (percentage of that cell population that registered greater than the baseline mean fluorescence intensity [expressed <1000 MFI]) of Dil+ peritoneal PMN and B1 cells from sham (normal saline)–pretreated versus poly I:C–pretreated, resiquimod-pretreated, and LPS-pretreated neonatal mice 24 hours after pretreatment and 2 hours after cecal slurry adminstration. (C) Percentage (percentage of PMNs that registered greater than the baseline mean fluorescence intensity [expressed <100 MFI]) of ROS production in PMNs from sham (normal saline)–pretreated versus poly I:C–pretreated, resiquimod-pretreated, and LPS-pretreated neonatal mice 24 hours after pretreatment and 2 hours after cecal slurry administration. (D) Percentage (percentage of PMNs that registered greater than the baseline mean fluorescence intensity [greater than 100 MFI]) of ROS production in PMNs from sham (normal saline)–pretreated versus TLR4(LPS)-pretreated neonatal mice 24 hours after pretreatment and measured 30 minutes, 2 hours, and 6 hours after cecal slurry administration. (E) Time course of recruitment of PMNs following TLR4 (LPS) pretreatment measured at 30 minutes, 2 hours, and 6 hours after cecal slurry as compared with sham pretreatment (normal saline). Data shown are representative of 3 separate experiments. Bar graphs represent means with standard deviation (*P < .05 compared with sham; ‡P < .05 as compared with sham and TLR7/8; †P < .05 as compared with all, by Student t test or 1-way ANOVA). In panels D and E, means are shown, with error bars representing SD.

Figure 7

TLR7/8 (resiquimod) and TLR4 (LPS) agonist–enhanced survival in neonatal mice with polymicrobial sepsis is independent of the adaptive immune system and type I IFN signaling. Kaplan-Meier survival curves for RAG-1–deficient (A) and IFNAR-null (B) neonates following sham (♦; normal saline) or TLR agonist pretreatment (resiquimod 5 μg/gm BW; ■) or LPS (1 μg/gm BW; ▵) versus wild-type sham (normal saline; ◇) and subsequent administration of cecal slurry (1.3 mg/g LD70).